Witnesses tell Senate committee U.S. advantage rests on workforce and diffusion, warn China faces diffusion deficit
Get AI-powered insights, summaries, and transcripts
Subscribe
Summary
Dr. Jeffrey Ding argued U.S. technological leadership depends on diffusion capacity — the ability to adopt AI across the economy — and presented evidence of a Chinese diffusion gap; witnesses called for workforce expansion and applied adoption programs.
Dr. Jeffrey Ding told the Senate Select Committee on Intelligence that assessing technological leadership requires distinguishing innovation capacity from diffusion capacity. "A state's ability to spread and adopt innovations after their initial introduction across productive processes" is central, Ding said, adding that AI is a general-purpose technology that produces economy-wide productivity effects when widely diffused.
Ding reported research findings that, by one baseline quality metric, China has 29 universities meeting an AI-engineering threshold compared with 59 in the United States; he said that gap exemplifies a diffusion deficit that limits China’s ability to turn model breakthroughs into broad economic or military advantage. He cautioned against overhyping Chinese model performance and urged policymakers to focus on adoption and workforce development.
Panelists and senators discussed policy levers to improve U.S. diffusion capacity: expand STEM education, broaden the base of applied AI engineers, create field service centers, use voucher programs to help small businesses adopt AI, and increase academic access to compute resources. Dr. Yann LeCun recommended government-supported computing infrastructure for noncommercial research and stronger industry–academia exchange programs.
Senators asked about outbound investment, export controls and whether U.S. capital flowing to foreign AI firms erodes competitive advantage. Ding recommended transparency measures to understand the scale and nature of outbound investment rather than treating all such flows as categorically harmful.
Committee members asked witnesses to return with specific legislative recommendations to widen the workforce and accelerate safe adoption of AI in priority sectors.